Long-term catheter-associated bacteriuria (LTCAB) is the most frequent infection in U.S. health care facilities. Although usually asymptomatic, stones, fever, acute pyelonephritis (pyelo), bacteremia, and death may result. This Program addresses two fundamental questions: 1) How do bacterial species colonize the catheterized urinary tract? 2) How do these bacteria damage or cross the epithelium? Three important research elements will be used: 1) Human renal proximal tubular epithelial cells (HRPTEC) from 98 donors; 2) LTCAB, pyelo, and fecal strains of E. coli and P. mirabilis, common in bacteremic UTI; 3) Non-catheterized and LTCAB mouse models of acute pyelonephritis. Each of four Projects uses state-of-the-art molecular methods: 1) E. coli Hemolysin Kills Human Renal Epithelial Cells. Of pyelo E. coli, 50% are hemolytic. Hemolytic e. coli kill HRPTEC; transposon mutagenesis demonstrated the cytotoxin to be hemolysin. This Project will develop 4 permutations of a prototypic pyelo E. coli strain (+\- hemolysin and +\- HRPTEC-adherence) and examine the 4 in the HRPTEC and mouse model systems. That HRPTEC from aged are more susceptible than those from young humans will be explored as will renal tropism of cytotoxicity and mechanism of hemolysin action. 2) Mechanisms of E. coli Invasion of the Renal Parenchyma. This Project will use in vitro human renal epithelial cell invasion models to study mechanisms of E. coli penetration of the epithelial barrier. E. coli causing different UTI syndromes will be compared for invasive ability, the effect of age on susceptibility to invasion will be explored, and transposon mutagenesis will be used to identify invasion factors. 3) Proteus mirabilis Pathogenesis of Urinary Tract Infection. P. mirabilis is a common isolate in LTCAB and is associated with stones, obstructed catheters, pyelo, and bacteremia. Genetic mutations have demonstrated its urease to be a potent virulence factor. Similar construction of mutants for MR/P fimbriae, flagella, and hemolysin will allow study of these as virulence factors in HRPTEC and mouse model systems. 4) Polymicrobial Colonization of Catheter Surfaces. Most patients with LTCAB have polymicrobial bacteriuria. 1st-time isolates of bacteriuric episodes will be screened in an in vitro catheter model for dominant species. Bacterial surface structures enhancing colonization, bacterial interaction with Tamm-Horsfall protein during colonization, and antibiotic- resistance of colonizing bacteria will be examined.